CT density of cervical thymus, in comparison with mediastinal thymus

Although the thymus is predominantly a mediastinal organ, part of it can often be seen in lower neck normally [1, 2]. When seen in the lower neck, cervical thymus has to be differentiated from pathologies like cervical lymph node, thyroid/parathyroid mass, or anatomic variants like enlarged distal thoracic duct [3]. In children, branchial cleft anomaly, teratoma, venous malformation, and lipoma also need to be excluded [4]. CT texture (lobules of soft tissue interspersed with fat) [1], similarity with CT density of mediastinal thymus [5] and continuity with mediastinal thymus on sagittal/coronal images, are given as the criteria for diagnosis of normal cervical thymus. The unique ultrasound appearance of thymic tissue (echogenic linear structures/foci in a hypoechoiec background) also allows a specific diagnosis of cervical thymus and pathology correlation is not indicated.

It is a common practice at our center to do USG correlation when there is any ambiguity in the neck (including cervical thymus) on CT scan. We encountered some cases where the cervical component of thymus had a density different from the mediastinal thymus, but ultrasound (along with CT continuity with mediastinal thymus on sagittal/coronal images) confirmed that the cervical lesion was indeed a normal variant and not some small lymph node/other pathology. We did a retrospective study of cases of the cervical thymus in which USG correlation was available, to compare CT densities of cervical and mediastinal parts of thymus.

The 22 studied patients included 10 males (45.5%) and 12 female patients (54.5%); age varied from 1 to 34 years (mean age 14); 19 patients had contrast-enhanced CTs and 3 patients had non-contrast CTs Additional file 1.

Embryological pathway of thymic descent is from the angle of mandible to the superior mediastinum (along thymopharyngeal duct); hence, thymus can be found anywhere along this path. Continuity with mediastinal thymus distinguishes cervical component of thymus from ectopic thymic tissue. Cervical component of thymus may be present at all ages [1], though it is more commonly imaged in children and young adults [2] (frequency decreases with age).

On ultrasound, thymus of infants may be homogeneous, with echogenicity slightly less than liver and spleen [6] or have multiple linear or branching echogenic foci [7]. In older children, the classical ‘starry sky’ pattern may be seen due to the mixture of lymphoid tissue and fat; it has also been called ‘speckled appearance’ due to the hyperechoiec septae and hyperechoiec foci on a hypoechoiec background [4]. Continuity with rest of the (mediastinal) thymus [4], and the same USG characteristics as mediastinal thymus [4] may be noted. Since the thymus is very pliable, cardiac pulsations and respiratory movements may affect its shape, and this may be noticeable on real-time ultrasound [5, 8]. The characteristic sonologic features of cervical thymus obviate the need for biopsy/surgical removal [4, 5, 9].

Thymic size and density are affected by a number of factors. Thymic size decreases with age; acute size reduction may occur during stress (infection, neoplasm, surgery, radiotherapy, steroid administration, and chemotherapy); true hyperplasia occurs in recovery from recent stress [10]; when reactive (true) hyperplasia exceeds 50% of original thymic size, it is called rebound hyperplasia. Fatty change is associated with increasing age, cigarette smoking, high BMI, and male sex [11]. Post-contrast CT density of > 41.2 HU was found to be suggestive of lymphofollicular hyperplasia (seen in a variety of immunologically mediated diseases) rather than true hyperplasia in one study [12], but further studies are needed.

Our study demonstrates that CT density of cervical part of thymus may vary from that of mediastinal part of the thymus, with density of cervical thymus being lower than the density of mediastinal component. But densities of cervical and mediastinal parts of thymus show positive correlation. We did not study the inter-observer variation in CT density, based on the excellent intra- and inter-observer agreements noted previously [11, 13, 14]. Decrease in CT density of thymus with age, noted previously in mediastinal thymus [11], has been demonstrated in the cervical thymus in our study.

The difference in CT density between cervical and mediastinal parts of the thymus may be partially due to the non-involvement of the cervical thymus in thymic hyperplasia. Most of our oncology patients would have had stress (neoplasm, infection, surgery, chemotherapy, radiotherapy, steroid administration) and reactive thymic hyperplasia after recovery from stress.

Limitations of our study include the retrospective nature of the study and absence of histological correlation. All patients in the study were imaged for non-uniform clinical reasons. Sick patients may have a higher rate of thymic hyperplasia and hence increased the difference in densities between cervical and mediastinal components of the thymus, but this is acceptable because it mirrors the actual incidence that would be seen on medical imaging.

CT densities of cervical and mediastinal components of thymus may vary, with density of cervical thymus being lower than the density of mediastinal component. When continuity with mediastinal thymus is seen on CT, diagnosis of the cervical thymus can be made even in the presence of different CT densities. Densities of cervical and mediastinal parts of thymus show positive correlation. Decrease in CT density of thymus with age, noted previously in the mediastinal thymus, has been demonstrated in cervical thymus also. Understanding these may help avoid confusion on CT and avoid the need for correlative USG, saving time and effort.